摘要：Objectives. We examined the association between regular cigarette smoking and new onset of mood and anxiety disorders. Methods. We used logistic regression analysis to detect associations between regular smoking and new-onset disorders during the 3-year follow-up among 34 653 participants in the longitudinal US National Epidemiologic Survey on Alcohol and Related Conditions (2001–2005). We used instrumental variable methods to assess the appropriateness of these models. Results. Regular smoking was associated with an increased risk of new onset of mood and anxiety disorders in multivariable analyses ( Fdf = 5,61 = 11.73; P < .001). Participants who smoked a larger number of cigarettes daily displayed a trend toward greater likelihood of new-onset disorders. Age moderated the association of smoking with most new-onset disorders. The association was mostly statistically significant and generally stronger in participants aged 18 to 49 years but was smaller and mostly nonsignificant in older adults. Conclusions. Our finding of a stronger association between regular cigarette smoking and increased risk of new-onset mood and anxiety disorders among younger adults suggest the need for vigorous antismoking campaigns and policy initiatives targeting this age group. The relationship between cigarette smoking and pulmonary and cardiovascular diseases is well established. 1–4 Indeed, early reports on the association of smoking and lung cancer were the major impetus for public health efforts to curb smoking. 5 The publication in 1964 of the surgeon general’s report on smoking and health was associated with a steady decline in smoking—a trend that in the following decades likely contributed to declining rates of lung cancer, especially among men. 6–8 The relationship between cigarette smoking and mental disorders, however, remains less well characterized. Several large community surveys have found increased prevalence of mental disorders among smokers. 9–18 Longitudinal studies have shown an increased risk of subsequent mood and anxiety disorders in individuals who smoked cigarettes and suggested that the effects might vary according to gender and age. 15,19–21 Although longitudinal studies can establish the temporal order of smoking and subsequent incidence of mental disorders, their findings can be confounded by factors that are not known or that are difficult to measure, such as a common genetic predisposition 22–25 or a specific personality diathesis, 26,27 and that could explain the observed association of smoking and mental disorders. Establishing the causal link between smoking and mental disorders conclusively would require experimental evidence, which is generally difficult or impossible to obtain. Exposing research participants to tobacco products in a randomized trial raises ethical issues. Controlled randomized trials of smoking cessation interventions might provide causal evidence. However, nicotine withdrawal is often associated with significant early psychological symptoms that may overshadow mental health benefits of smoking cessation. 28 Furthermore, randomized trials often comprise a highly selected group of participants, and their findings may not be generalizable. Because of these limitations, prospective observational studies of representative population-based samples may provide the best available means to investigate the causal links between smoking and common mood and anxiety disorders. Confidence in causal inferences from such data may be enhanced by the use of statistical methods designed for such inference, such as instrumental variable methods. 29–31 In this study, we examined the association between smoking and subsequent development of mood or anxiety disorders in data from the longitudinal National Epidemiologic Survey on Alcohol and Related Conditions (NESARC), a large prospective epidemiological survey of the US population. 32–34 We examined the association of regular smoking during at least part of the 3-year follow-up period with new onset of major depressive episodes, dysthymia, manic episodes, generalized anxiety disorder, panic disorder, social phobia, specific phobias, and posttraumatic stress disorder. Our hypothesis, derived from previous research, 9,10,12,20,35,36 was that regular smoking is associated with an increased risk of new onset of these mood and anxiety disorders. We also assessed the dose–response relationship between the average number of cigarettes smoked and the risk of new-onset disorders and whether the relationship between smoking and new-onset psychopathology varied by sociodemographic characteristics.
